case study digital transformation and artificial intelligence topic teachable machine application for waste classification

UNIVERSITY OF ECONOMICS AND LAW FACULTY OF INFORMATION SYSTEMS CASE STUDY DIGITAL TRANSFORMATION AND ARTIFICIAL INTELLIGENCE Topic: TEACHABLE MACHINE APPLICATION FOR WASTE CLASSIFIC

UNIVERSITY OF ECONOMICS AND LAW

FACULTY OF INFORMATION SYSTEMS

CASE STUDY DIGITAL TRANSFORMATION AND ARTIFICIAL INTELLIGENCE

Topic: TEACHABLE MACHINE APPLICATION FOR WASTE

CLASSIFICATION Lecturer: Nguyen The Dai Nghia Group 3:

1 MSSV: K214060444 — Hoang Trung

2 MSSV: K214060436 — Nguyen Yen Nhi

3 MSSV: K214061755 — Le Thi Cam Hang

4 MSSV: K214060445 — Nguyen Thanh Minh Tu

Ho Chi Minh City, December 2022

CONTENTS

TT Number 2á an 3

“Ji (¡i22 .(1AẢÓẢÍ 3

` ng 4

3S Or Gamic WASKEL n6 ‹(44 8

h0 a-.ẦẢdẢa 9 Mi) 1 10

2 Sample task and analysis report including reminder deadline: 10

Project Overview:

Name of Project: Teachable Machine Application For Waste Classification Link of Project: https://teachablemachine withgoogle.com/models/o-Kwdw0A8/ Environmental issues are always a "hot" issue of concern to everyone People's living environment is currently facing serious problems because of production activities and human activities, in which the problem of waste is one of the main problems causing the environment to become more andintending to help all types of garbage cause environmental pollution, and some types of waste, if properly treated and classified, can reduce the pollution level in the environment That's why our team undertook this project with the aim of helping to separate different types of waste so that we can use different treatment methods to improve the quality of the environment

Il Number of class:

The above project uses 3 classes The first class 1s the medical waste class that includes used medical masks, used medical gloves and used medical needles The second class contains recyclable waste like olclassesstic cups and used water bottles The next class is organic waste consisting of whiche peels, leaves and flowers And the forth class, is non-recyclable inorganic waste that includes broken glass and sharp objects can no longer be used Finally, the last class is others that includes face, human, smartphone and background

lil Functions and applicability of the Project:

1 Functions:

With this project, we will teach AI how to differentiate between five types of waste: medical waste, organic waste, non-recyclable inorganic waste, recyclable waste, and other waste, allowing us to determine which type of waste requires which treatment method For example, recyclable waste such as plastic bottles or styrofoam boxes can

be used as potted plants to grow omamental plants in the family while helping to reduce the amount of plastic waste discharged into the environment every day It helps

to make our living environment greener

2 Applicability:

As analyzed above, the application of the project is quite potential because it can be applied to the classification of garbage, and almost every industry currently has waste, especially in the medical or other industries contact with chemicals or special items that require special disposal This AI can be used in people's daily lives to help them avoid headaches and confusion about when to throw trash in which bin In addition, this AI can also teach children how to distinguish the above types of garbage to educate them early on the classification of garbage, which can help them improve their awareness about protection early living environment

IV Project Evaluate:

In terms of the project, this is the first project we have worked on with a teachable machine, so mistakes are inevitable The first is that it can't completely separate all types of waste because there are some types of waste that we can't find, such as chemicals, specialized items from medical facilities, or some types of garbage The second is that sometimes the machine will recognize the wrong type of garbage that is put on the screen The team is still working to correct the error But in general, the group's AI has distinguished the things that the group teaches the AI, in which there will be some errors but most of them will be recognized by the AI Regarding future projects, the team will try to collect all the samples to be able to come up with the most complete product

V Steps:

1 Step 1:

Go to the website https://teachablemachine.withgoogle.com/ and you will have an interface as follows:

2 Step 2:

Click Get Started and you will get a screen like the following

3 Step 3:

Click on Image Project then click on the standard image model

4, Step 4:

After clicking, a window will appear for you to teach AI how to distinguish according to each class you choose:

5 Step 5:

You can choose to teach the AI by uploading the product or by adding a picture of the product

6 Step 6:

After completing the input of the data, I click the trained model to teach the AI how

to distinguish the above data

Training

Train Model

VI Product Testing:

1 Medical waste:

Preview this model live

How does it work?

Machine learning models are trained on examples (e.g images,

sounds, poses) gathered by the creator Thelr results depend on the

data they've been trained on

Want to use this model in your project?

See this link to learn how to use Teachable Machine models in your

projects

Report this model: Output

If you have concerns about this model, report it using this form

yté

thể tải

{C] teachablemachine.withgoogle.com/models/Ne5lqYZAz/ chế

Khác

[3 modeljson

Í matadata ison

2 Recyclable waste:

3 Organic waste:

4 Non-recyclable inorganic waste:

5, Others:

VII Trello:

1 Assign tasks to members:

10

2 Sample task and analysis report including reminder deadline:

3 Worksheet form when completed tasks:

End

lãi